Extended TAILOR-PCI Data Again Show No Gains for Genotype-Guided Meds
Believing still in some benefit of testing, Naveen Pereira said the P value shouldn’t be viewed as “binary.”
A genotype-guided strategy for deciding on antiplatelet agents after PCI does not lead to improvements in long-term ischemic events, according to extended follow-up from TAILOR-PCI.
The results are in line with what was seen in the original 12-month findings released last year, when TAILOR-PCI failed to show that using a more-potent P2Y12 inhibitor (ticagrelor) instead of clopidogrel in patients with a loss-of-function allele who had undergone PCI for ACS or stable CAD reduced the risk of ischemic events at 1 year. Notably, TAILOR-PCI did demonstrate a benefit of genotype-based treatment at an earlier follow-up point—3 months—and in reducing multiple ischemic events per patient.
Investigators had hoped, however, that as more events accumulated in both arms over time, a significant difference would emerge. The trial was initially designed to show a 50% difference in events between groups but fell short, with only a 34% difference that missed statistical significance.
Presenting their analysis—which included a median follow-up of 39 months—at the recent American College of Cardiology (ACC) 2021 Scientific Session, Naveen Pereira, MD (Mayo Clinic, Rochester, MN), stressed that the findings, while not what investigators had hoped, need to be viewed with caution. Patients in the genotype-guided group, he noted, were less likely to continue their P2Y12 inhibitor after 1 year, especially those on ticagrelor, compared with those in the conventional group.
“Interpretation of these results should be made in the context that we did not mandate what antiplatelet therapy and who should be on antiplatelet therapy after 12 months,” he said. “Hence, we observed a very variable P2Y12 inhibitor use beyond 12 months to the detriment of the genotype-guided group where there was a sharp drop-off in the use of ticagrelor. But there was continued use of clopidogrel in the majority of the patients in the conventional group.”
While the study thus failed to overcome some of the limitations seen with the earlier 12-month findings, said Pereira, “I don't think it should dissuade physicians or medical care providers from doing genetic testing because this is an observational study.”
Judith S. Hochman, MD (NYU School of Medicine, New York, NY), who served as a discussant for the study, said the analysis “certainly importantly contributes to the evidence base around a very controversial question: to use testing-guided therapy or not for patients who undergo PCI in terms of selection of the P2Y12 inhibitor.”
For Hochman, the results are “very clear,” providing a firm “no” to the question of whether genotype-guided antiplatelet decision making reduced ischemic outcomes over the long-term, even if there was an earlier signal of benefit.
Still No Significant Benefit
For the study, Pereira and colleagues followed 4,747 (including 1,700 loss-of-function carriers) of the original 5,276 patients from TAILOR-PCI who were randomized to receive genotype-guided (ticagrelor 90 mg BID for loss-of-function carriers and clopidogrel 75 mg daily for noncarriers) or conventional clopidogrel therapy (75 mg daily) following PCI between May 2013 and October 2018.
Notably, the percentage of person-days on dual antiplatelet therapy dropped in the genotype-guided and conventional therapy arms from 98% to 43% and 57%, respectively, over the long term. Ticagrelor use dropped as well in the study arm, with only 23% of person-days reported at follow-up compared with 75% at 1 year.
There was no difference in the primary endpoint—composite of cardiovascular death, MI, stroke, stent thrombosis, and severe recurrent ischemia—between patients who were treated according to a genotype-guided or conventional strategy over extended follow-up (adjusted HR 0.95; 95% CI 0.70-1.29). There was also no difference in the safety endpoint of TIMI major or minor bleeding (adjusted HR 1.10; 95% CI 0.60-2.04).
Results were maintained in several subgroup analyses.
“The benefit of genotype-guided therapy and the effect of the drug-gene interaction appears to be most evident within the first 3 months after PCI,” Pereira said. “So you really have to be on the drug and have the loss-of-function genotype to see the beneficial effect, and it may so happen that the most vulnerable patients essentially were flushed out in the first 3 months because this is the time to first event type of analysis.”
Hochman said she was struck by how many patients stopped ticagrelor over the long term, suggesting maybe increased cost and side effects played a role.
Early on, she continued “there was a lot of controversy” over the results as the P value for the 12-month primary endpoint was 0.056, a technically nonsignificant result. “I would contend that we've learned a lot about Bayesian analyses and we don't live and die based on a P value of 0.05—that you don't think [with] 0.049 there's is a difference and 0.051 there's not.”
Pereira agreed, saying that the P value should not be viewed in such a “binary fashion.” He reported that the probability of TAILOR PCI being a negative trial using a Bayesian calculation is 2.9%. “I think that the effect size and the tightness of the confidence intervals really point to that fact,” he added.
Currently, the ACC and American Heart Association don't endorse a testing-guided strategy for P2Y12 inhibitor selection, Hochman noted, adding that recent meta-analyses including one from April show “that either a platelet-function-testing strategy or genotype-testing strategy is superior to not testing; however, the absolute reduction in events is small. And you just showed us that over the long term, there's no difference in events.”
Given the known associations of race and certain STEMI presentations benefitting from longer-term therapy, she asked: “Do we need more precision medicine to select a certain risk group to actually do the testing in?”
“Taking other clinical variables in addition to genotype and perhaps coming up with a score to predict the high-risk patient could be useful to become more precise with this precision-medicine approach,” Pereira said. “Our study points to the fact that we could have an intermediate step right now that could change our practice. Find out what the genotype of the patient is; if it's noncarrier patients, you can safely give them clopidogrel, and if they are loss of function, give them ticagrelor or prasugrel.”
The only way to undoubtedly answer the question of whether a genotype-guided strategy improves long-term outcomes would be “to do a DAPT-like trial where patients are on protocol-guided treatments beyond 12 months,” Pereira said, adding that this will be unlikely due to the large size of the study that would be needed.
For now, “I believe the signal for genetic guidance in helping our patients is definitely there,” he added. But for it to be embedded in clinical practice, Pereira said, the tests will have to be easily accessible for clinicians and affordable.
Pereira N. Effect of genotype-guided oral P2Y12 inhibitors vs conventional clopidogrel on long-term ischemic outcomes after percutaneous coronary intervention: the TAILOR-PCI randomized clinical trial follow-up study. Presented at: ACC 2021. May 17, 2021.
- The study was funded by the National Heart, Lung, and Blood Institute.
- Pereira reports no relevant conflicts of interest.